Differential Loneliness Scale (DLS)

Abstract

The Differential Loneliness Scale (DLS), developed by psychometric researchers Neena Schmidt and Vello Sermat in 1983, represents a foundational multidimensional assessment instrument designed to measure subjective feelings of loneliness across distinct relational domains. Departing from prevailing unidimensional conceptualizations of loneliness—such as the widely utilized UCLA Loneliness Scale—the DLS posits that loneliness is not a global, undifferentiated affective state, but rather a differentiated phenomenon arising from perceived deficits within specific qualitative spheres of an individual’s interpersonal network. The full adult version of the instrument comprises 60 self-report items evaluated through a dichotomous (True/False) response format. The scale systematically operationalizes loneliness across four primary relational subscales: (1) Romantic-Sexual Relationships (12 items), (2) Friendships (22 items), (3) Family Relationships (18 items), and (4) Relationships with the Community or Larger Social Groups (8 items).

Psychometric evaluations across non-clinical, student, and clinical populations demonstrate robust psychometric properties. The DLS exhibits high internal consistency, with overall scale reliability estimates consistently exceeding Kuder-Richardson Formula 20 (KR-20) and Cronbach’s alpha coefficients of .90 to .92, alongside subscale reliabilities ranging from .75 to .89. Test-retest reliability across multi-week intervals demonstrates adequate temporal stability (coefficients ranging between .85 and .91). Extensive construct, convergent, and discriminant validity investigations reveal meaningful correlations with global loneliness, depressive symptom inventories, social anxiety, attachment insecurity, and interpersonal alienation scales, while retaining distinct divergent variance that underscores the functional independence of each relational sphere. The DLS remains a critical assessment tool in clinical psychology, social psychology, marital and family counseling, and organizational behavior for diagnosing targeted social deficits and evaluating tailored interpersonal interventions.

Keywords

Differential Loneliness Scale, DLS, multidimensional loneliness, relational deficits, social isolation, emotional loneliness, psychometrics, friendship satisfaction, family cohesion, romantic relationship loneliness

Authors

The Differential Loneliness Scale was formulated and psychometrically validated by Neena Schmidt, Ph.D., and Vello Sermat, Ph.D., operating within the Department of Psychology at York University in Toronto, Ontario, Canada. Dr. Vello Sermat was a distinguished social psychologist recognized for his pioneering experimental and empirical investigations into human conflict resolution, interpersonal communication, trust dynamics, and relational deprivation. Dr. Neena Schmidt collaborated extensively on the structural decomposition of relational dissatisfaction, focusing specifically on psychometric test construction and the operationalization of relational deficits.

Historically, the instrument was conceptualized and refined throughout the late 1970s and early 1980s, culminating in their seminal 1983 monograph published in the Journal of Personality and Social Psychology. Both researchers sought to resolve existing methodological ambiguities in clinical assessment where clients presented with pervasive feelings of alienation despite possessing intact, objective social networks in specific life sectors. Inquiries regarding archival access and psychometric materials have traditionally been maintained via the academic archives of York University and research repositories focusing on interpersonal health.

Purpose

The primary purpose of the Differential Loneliness Scale is to provide an empirically grounded, multidimensional diagnostic and research tool capable of differentiating loneliness across diverse interpersonal systems. Prior to the development of the DLS, psychological research largely treated loneliness as a monolithic state. Instruments developed during that era aggregated an individual’s overall sense of isolation without identifying which specific social relationships failed to meet their fundamental psychological needs. Consequently, an individual who experienced profound romantic isolation but maintained a robust network of supportive friends might obtain an intermediate score on a unidimensional measure, obscuring the clinical reality of their distress.

The DLS addresses this structural limitation by mapping social deficits directly onto ecological relationship domains. From a theoretical perspective, the scale is designed to examine the hypothesis that interpersonal deficits in one life domain (e.g., romantic-sexual partnerships) cannot be completely ameliorated or substituted by surpluses in another domain (e.g., strong familial connections). Clinically, the instrument provides therapists, marriage and family counselors, and clinical psychologists with a targeted diagnostic profile. By examining subscale elevations, clinicians can pinpoint whether therapeutic interventions should focus on couple-level intimacy, peer group assimilation, family-of-origin conflict resolution, or broader community integration.

In empirical research, the DLS serves several key functions:

• Epidemiological and Demographic Profiling: It enables researchers to trace how specific types of loneliness shift across the lifespan—such as the transition from family-centric loneliness in late adolescence to romantic loneliness in emerging adulthood, or community isolation among older adults.
• Evaluating Interpersonal Interventions: The DLS functions as an outcome measure in randomized controlled trials of group psychotherapy, social skills training, and systemic family therapies, allowing investigators to detect domain-specific treatment gains.
• Organizational and Occupational Assessment: As demonstrated by researchers evaluating workplace isolation, the DLS provides a foundation for assessing how occupational displacement, remote work arrangements, and corporate restructuring disrupt an individual’s connection to broader communal and peer support structures.

Psychological Construct

The core psychological construct measured by the Differential Loneliness Scale is relational loneliness, defined as an unpleasant emotional, cognitive, and affective reaction to a perceived discrepancy between an individual’s desired and achieved interpersonal relationships within specific domains. The DLS operationalizes this construct across four distinct, correlated yet functionally separable dimensions:

1. Romantic-Sexual Relationships

This subscale (12 items) assesses feelings of deprivation, alienation, and dissatisfaction stemming from the absence or qualitative inadequacy of an intimate romantic partnership, marriage, or committed love relationship. The items measure subjective appraisals of reciprocal emotional investment, mutual cooperation, sexual fulfillment, and profound emotional understanding between partners. Representative items measure whether the respondent possesses a partner who fulfills their emotional security needs (e.g., Item 21) or whether they perceive emotional vulnerability as dangerous or unreciprocated (e.g., Item 35). A high score reflects romantic loneliness, emotional isolation, and the absence of a primary attachment figure capable of offering unconditional validation.

2. Friendships

The friendship subscale (22 items) evaluates perceived deficits within the individual’s peer network and platonic social circle. The items assess peer availability, emotional disclosure, dependability, mutual trust, and social companionship. Rather than simply quantifying the absolute number of associates, this dimension examines whether friends can be relied upon during moments of acute crisis (e.g., Item 11), whether friendships result in frequent interpersonal disappointment (e.g., Item 30), and whether the individual experiences social anxiety or behavioral inhibition when initiating social activities (e.g., Item 53). Deficits in this domain indicate peer alienation, lack of a reliable peer alliance, and a subjective sense of unreciprocated social investment.

3. Family Relationships

Comprising 18 items, this dimension examines relational quality, emotional security, belongingness, and psychological validation within the individual’s family of origin or primary kin network. The items assess whether the respondent experiences positive emotional exchange with relatives (e.g., Item 1), whether their voice is respected in collective decision-making processes (e.g., Item 59), and whether family dynamics are characterized by critical hostility, estrangement, or social withdrawal (e.g., Item 44 and Item 57). Deficits in this sphere capture familial alienation, parental/sibling detachment, and an unfulfilled need for primary kinship belonging.

4. Relationships with the Community or Larger Social Groups

This subscale (8 items) captures the individual’s integration into their broader ecological environment, neighborhood, voluntary associations, and cultural or civic communities. Rather than focusing on dyadic intimacy, this dimension addresses macro-level social integration, collective solidarity, and the availability of secondary institutional support systems during times of distress (e.g., Item 8). Items measure feelings of anonymity among co-located individuals (e.g., Item 2), perceived neighborhood apathy (e.g., Item 27), and social validation within group settings (e.g., Item 56). High scores on this subscale reflect societal alienation, collective anomie, and an absence of communal belonging.

Theoretical Framework

The theoretical architecture of the Differential Loneliness Scale is directly rooted in Robert S. Weiss’s (1973) seminal typology of loneliness, further integrated with the cognitive discrepancy model formulated by Letitia Anne Peplau and Daniel Perlman (1982), and John Bowlby’s attachment theory.

Weiss’s Relational Provision Model

Robert Weiss posited that loneliness is not a unitary state, but is fundamentally partitioned into two distinct typologies based on specific relational provisions: the loneliness of emotional isolation and the loneliness of social isolation. Emotional isolation arises from the absence of a close, primary emotional attachment—traditionally provided by a spouse, romantic partner, or intimate confidant—and is characterized by feelings of desolation, subjective apprehension, and vigilance. In contrast, social isolation results from the lack of an engaging social network, peer group, or community membership, yielding feelings of boredom, marginalization, and aimlessness. Schmidt and Sermat extended Weiss’s dual model by bifurcating emotional isolation into romantic/sexual and familial dimensions, while simultaneously dividing social isolation into peer-group (friendship) and macro-social (community) components.

The Cognitive Discrepancy Model

The DLS is theoretically anchored in the cognitive framework of Peplau and Perlman, which defines loneliness as the cognitive perception of a mismatch between one’s internal interpersonal standards and the reality of one’s existing social connections. Under this model, objective social isolation (such as living alone or having a low frequency of daily interactions) does not inevitably generate loneliness. Rather, loneliness manifests when an individual cognitively appraises their existing relationships as failing to meet their expectations for intimacy, trust, or mutual aid. The items across all four subscales of the DLS explicitly operationalize these subjective cognitive appraisals rather than tallying objective interaction counts.

Attachment and Social Exchange Foundations

Attachment theory informs the structural stability of the DLS subscales. Internal working models of self and others, formed during early attachment interactions, dictate an individual’s capacity to navigate adult romantic intimacies (Romantic subscale) and maintain stable kinship ties (Family subscale). Concurrently, social exchange theory provides the rationale for the Friendship and Community subscales, wherein relationships are sustained via reciprocal exchanges of emotional support, tangible resources, and behavioral validation. When an individual perceives persistent imbalances in interpersonal reciprocity (e.g., feeling that friends take advantage or that community members are indifferent), subjective loneliness inevitably escalates.

Validity

Extensive psychometric investigations have established strong empirical support for the construct, convergent, discriminant, and predictive validity of the Differential Loneliness Scale.

Construct and Factorial Validity

During the initial scale development by Schmidt and Sermat (1983), the structural independence of the four domains was established across multiple samples consisting of university undergraduates, community adults, and clinical outpatients. Confirmatory investigations confirmed that a four-factor oblique model provided a significantly superior fit to empirical data compared to a unidimensional model, demonstrating that the four subscales measure distinct relational entities that cannot be collapsed into an undifferentiated composite without substantial loss of diagnostic precision.

Convergent Validity

Convergent validity has been established through systematic correlations with external measures of interpersonal distress, subjective well-being, and social support:

• Global Loneliness: The DLS total score correlates strongly with the revised UCLA Loneliness Scale, with correlation coefficients typically ranging between r = .68 and r = .78 (p < .001). Furthermore, specific subscales demonstrate logical convergent patterns; for instance, the Friendship and Community subscales exhibit the highest correlations with UCLA items targeting peer integration.
• Depressive Symptomatology: The DLS demonstrates moderate to strong positive correlations with the Beck Depression Inventory (BDI) (r values typically between .42 and .56), reflecting the established clinical comorbidity between interpersonal alienation and dysphoric mood.
• Social Support Inventories: Statistically significant inverse correlations emerge between the DLS subscales and established social support measures, such as the Social Support Questionnaire (SSQ) and the Multidimensional Scale of Perceived Social Support (MSPSS). The Family and Friendship subscales correlate negatively with their corresponding domain scores on the MSPSS (r = -.55 to -.67).

Discriminant Validity

Discriminant validity analyses confirm that the DLS captures loneliness rather than overlapping personality traits or general negative emotionality:

• Introversion vs. Neuroticism: While the DLS correlates modestly with neuroticism on the Eysenck Personality Inventory (EPI) (r = .30 to .38), it remains clearly distinct from trait introversion. Highly introverted individuals who possess qualitative satisfaction within their chosen intimate relationships do not exhibit elevated scores on the DLS.
• Social Desirability: Correlations between the DLS subscales and the Marlowe-Crowne Social Desirability Scale are uniformly low to negligible (r = -.08 to -.16), demonstrating that respondent scoring is not heavily biased by social presentation tendencies.

Predictive and Criterion Validity

The scale effectively predicts behavioral and situational markers of relational disruption. Schmidt and Sermat demonstrated that individuals experiencing acute marital separation, divorce, or bereavement exhibited marked elevations specifically on the Romantic-Sexual Relationships subscale, while displaying normal distributions on the Friendship and Family subscales. Conversely, individuals relocating to new cities or institutional environments showed primary elevations on the Friendship and Community subscales. In workplace literature, such as research conducted by Marshall, Michaels, and Mulki (2007), adaptations of the community and peer dimensions accurately predicted job turnover intentions, organizational isolation, and burnout.

Reliability

The Differential Loneliness Scale exhibits excellent reliability across diverse cultural, clinical, and community cohorts, characterized by high internal consistency and consistent temporal stability.

Internal Consistency

Because the DLS utilizes a dichotomous (True/False) response format, internal consistency was originally calculated using the Kuder-Richardson Formula 20 (KR-20), alongside standard Cronbach’s alpha estimations. Across the normative validation samples:

• Total DLS Score: Yields an internal consistency reliability coefficient between α = .90 and α = .92, demonstrating exceptional overall reliability.
• Romantic-Sexual Relationships Subscale (12 items): Demonstrates internal consistency ranging from α = .86 to .89.
• Friendships Subscale (22 items): Demonstrates high internal consistency ranging between α = .83 and .88.
• Family Relationships Subscale (18 items): Yields coefficients consistently between α = .85 and .89.
• Relationships with Community Subscale (8 items): Displays acceptable to strong reliability for a brief 8-item scale, yielding coefficients from α = .74 to .80.

Test-Retest Stability

Evaluation of temporal stability across non-intervention intervals supports the scale’s reliability. Over a 2-week test-retest period, Schmidt and Sermat reported an aggregate stability coefficient of r = .91 (p < .001). Over extended intervals of 6 to 8 weeks, test-retest coefficients remained robust (ranging from r = .79 to .85 across subscales). Importantly, longitudinal research demonstrates that while the subscale scores remain stable under constant environmental conditions, they show sensitive, statistically significant shifts in response to major life transitions (e.g., divorce, therapeutic intervention, initiation of a new romantic partnership), confirming both stability and responsiveness.

Factor Analysis

The structural composition of the Differential Loneliness Scale has been rigorously evaluated via both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

Exploratory Factor Analysis (EFA)

In the original instrument derivation, an initial pool of over 100 candidate items was subjected to principal components analysis followed by orthogonal (Varimax) and oblique (Promax) rotations. Four primary factors consistently emerged with eigenvalues greater than 1.5, collectively accounting for the majority of the common variance. Items demonstrating low primary factor loadings (< .40) or substantial cross-loadings (> .30 on secondary factors) were systematically eliminated, resulting in the final 60-item instrument.

The empirical factor structure directly validated Schmidt and Sermat’s four hypothesized relational domains:

• Factor 1 (Friendships): Loaded heavily with items referencing peer reliability, social self-efficacy, and casual recreational companionship (e.g., Items 4, 11, 22, 24). Factor loadings ranged from .42 to .71.
• Factor 2 (Family Relationships): Defined by items measuring kinship acceptance, communication transparency, and mutual familial respect (e.g., Items 1, 14, 20, 36). Factor loadings ranged from .45 to .76.
• Factor 3 (Romantic-Sexual Relationships): Captured items addressing dyadic exclusivity, partner emotional responsiveness, and intimacy (e.g., Items 9, 18, 21, 31). Factor loadings ranged from .50 to .82.
• Factor 4 (Community / Social Groups): Clustered items concerning neighborhood support, civic belonging, and group participation (e.g., Items 8, 12, 27, 49). Factor loadings ranged from .41 to .68.

Confirmatory Factor Analysis (CFA)

Subsequent psychometric investigations have subjected the 60-item four-factor structure to confirmatory factor analysis. Goodness-of-fit evaluations confirm that the four-factor oblique model demonstrates acceptable to strong fit indices across diverse demographic samples:

• Comparative Fit Index (CFI): Ranging between .90 and .94 across validation studies.
• Tucker-Lewis Index (TLI): Typically exceeding .90.
• Root Mean Square Error of Approximation (RMSEA): Estimates consistently fall between .042 and .055 (90% CI [.038, .060]), indicating excellent approximation and control for parsimony.
• Standardized Root Mean Square Residual (SRMR): Observed values consistently remain below .06.

Inter-factor correlations in CFA studies typically range from moderate (r = .28 between Romantic and Community) to moderately high (r = .52 between Friends and Community), confirming that while these domains share general variance attributable to an individual’s overall social functioning, they represent conceptually distinct interpersonal arenas.

Instrument / Measurement Tool

The Differential Loneliness Scale is a comprehensive psychometric instrument engineered for rapid completion and objective scoring. Below is a structured summary of the instrument’s operational characteristics:

• Test Type: Multidimensional self-report psychological assessment / inventory.
• Target Population: Adults and adolescents aged 16 and older; applicable in clinical, counseling, organizational, and general population research.
• Item Count: 60 items.
• Response Format: Forced-choice dichotomous format (TRUE or FALSE).
• Administration Time: Approximately 10 to 15 minutes.
• Subscale Breakdown:
  • Romantic/Sexual Relationships (12 items): Items 5, 9, 15, 18, 21, 28, 31, 35, 41, 45, 52, 58.
  • Friendships (22 items): Items 3, 4, 7, 11, 13, 16, 19, 22, 23, 24, 26, 30, 33, 37, 39, 43, 46, 48, 51, 53, 55, 60.
  • Family Relationships (18 items): Items 1, 6, 10, 14, 20, 25, 29, 32, 34, 36, 40, 42, 44, 47, 50, 54, 57, 59.
  • Relationships with the Community (8 items): Items 2, 8, 12, 17, 27, 38, 49, 56.
• Scoring Protocol:
  • Each item is scored either 0 or 1 point toward the loneliness construct.
  • Direct-Scored Items (No asterisk): Items describing feelings of isolation, dissatisfaction, or interpersonal deficits. A marking of TRUE (T) = 1 point; FALSE (F) = 0 points.
  • Reverse-Scored Items (Marked with an asterisk *): Items describing healthy, satisfying, supportive, or positive interpersonal experiences. A marking of FALSE (F) = 1 point; TRUE (T) = 0 points.
  • Total Score Range: 0 to 60. Higher total scores denote greater global loneliness.
  • Subscale Score Ranges:
    • Romantic/Sexual: 0 to 12
    • Friendships: 0 to 22
    • Family Relationships: 0 to 18
    • Community Relationships: 0 to 8

Permissions & Fee and Test Year

The Differential Loneliness Scale was published in 1983 by Neena Schmidt and Vello Sermat through the American Psychological Association (APA). The foundational empirical validation study appeared in the Journal of Personality and Social Psychology (Vol. 44, Iss. 5, pp. 1038–1047).

Licensing and Research Use: In accordance with standard academic dissemination guidelines, the scale items were published directly within academic literature and publicly deposited across research portals (including York University psychometric archives and the Fetzer Institute’s interpersonal instrument repository). The DLS is generally available free of charge for non-commercial, academic, scientific, and educational research purposes, provided that appropriate bibliographic credit is accorded to the original authors. For commercial applications, diagnostic software integration, or clinical corporate deployments, permission should be coordinated through the original copyright holder (American Psychological Association) or the institutional representatives of the primary authors.

References

The following peer-reviewed publications document the empirical development, psychometric properties, and clinical applications of the Differential Loneliness Scale and related conceptual models:

• Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). Basic Books.
• Marshall, G. W., Michaels, C. E., & Mulki, J. P. (2007). Workplace isolation: Exploring the construct and its measurement. Psychology & Marketing, 24(3), 195–223. https://doi.org/10.1002/mar.20158
• Peplau, L. A., & Perlman, D. (Eds.). (1982). Loneliness: A sourcebook of current theory, research and therapy. John Wiley & Sons.
• Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472–480. https://doi.org/10.1037/0022-3514.39.3.472
• Schmidt, N., & Sermat, V. (1983). Measuring loneliness in different relationships. Journal of Personality and Social Psychology, 44(5), 1038–1047. https://doi.org/10.1037/0022-3514.44.5.1038
• Weiss, R. S. (1973). Loneliness: The experience of emotional and social isolation. The MIT Press.

Items of the Scale